The expectation on older kernels (< 3.19) was to have the network
namespace always be a directory. This is not true if the network
namespace is bind mounted to a file, and will make the plugin fail
erroneously in such cases. The fix is to remove this assumption
completely and just do a basic check on the file system types being
returned.
Fixes #288
The veth is moved from the container NS to the host NS.
This is handled by the code that sets the link to UP but the wrong
hostVeth is returned to the calling code.
If interface name for a container provided by a user is already present,
Veth creation fails with incorrect error.
If os.IsExist error is returned by makeVethPair:
* Check for peer name, if exists generate another random peer name,
* else, IsExist error is due to container interface present, return error.
Fixes #155
This will give deterministic MAC addresses for all interfaces CNI
creates and manages the IP for:
* bridge: container veth and host bridge
* macvlan: container veth
* ptp: container veth and host veth
* _suite.go and _test.go file should be in the same package, using the
_test package for that, which requires some fields and methods to be
exported
* Introduce error type for cleaner error handling
* test adaptions for error type checking
Previously, the log lines appeared in stdout before the JSON encoding of
the error message. That would break JSON parsing of stdout. Instead, we use
stderr for these unstructured logs, consistent with the CNI spec.
Allow strings to be unmarshalled for CNI_ARGS
CNI_ARGS uses types.LoadArgs to populate a struct.
The fields in the struct must meet the TextUnmarshaler interface.
This code adds a UnmarshallableString type to assist with this.
Add a namespace object interface for somewhat cleaner code when
creating and switching between network namespaces. All created
namespaces are now mounted in /var/run/netns to ensure they
have persistent inodes and paths that can be passed around
between plugin components without relying on the current namespace
being correct.
Also remove the thread-locking arguments from the ns package
per https://github.com/appc/cni/issues/183 by doing all the namespace
changes in a separate goroutine that locks/unlocks itself, instead of
the caller having to track OS thread locking.
Previously this code used a run-time map lookup keyed by
runtime.GOOS/GOARCH. This version uses conditional compilation to make
this choice at compile time, giving immediate feedback for unsupported
platforms.
/proc/self/ns/net gives the main thread's namespace, not necessarily
the namespace of the thread that's running the testcases. This causes
sporadic failures of the tests.
For example, with a testcase reading inodes after switching netns:
/proc/27686/task/27689/ns/net 4026532565
/proc/self/ns/net 4026531969
/proc/27686/task/27689/ns/net 4026532565
See also:
008d17ae00
Running Suite: pkg/ns Suite
===========================
Random Seed: 1459953577
Will run 6 of 6 specs
• Failure [0.028 seconds]
Linux namespace operations
/cni/gopath/src/github.com/appc/cni/pkg/ns/ns_test.go:167
WithNetNS
/cni/gopath/src/github.com/appc/cni/pkg/ns/ns_test.go:166
executes the callback within the target network namespace [It]
/cni/gopath/src/github.com/appc/cni/pkg/ns/ns_test.go:97
Expected
<uint64>: 4026531969
to equal
<uint64>: 4026532565
/cni/gopath/src/github.com/appc/cni/pkg/ns/ns_test.go:96
------------------------------
•••••
Summarizing 1 Failure:
[Fail] Linux namespace operations WithNetNS [It] executes the callback within the target network namespace
/cni/gopath/src/github.com/appc/cni/pkg/ns/ns_test.go:96
Ran 6 of 6 Specs in 0.564 seconds
FAIL! -- 5 Passed | 1 Failed | 0 Pending | 0 Skipped --- FAIL: TestNs (0.56s)
FAIL
